J. Ph_ys. Chem. SoZids
Pergamon
ASSOCIATION
Press 1960. Vol. 17, Nos. l/2, pp. 69-79.
AND
DISSOCIATION
LUMINESCENT H. Philips Research Laboratories, (Received
OF CENTRES
IN
ZnS-In
KOELMANS
N. V. Philips’ Gloeilampenfabrieken, 24 September
Printed in Great Britain.
1959; revised 14 January
Eindhoven,
Netherlands
1960)
Abstract-ZnS-In phosphors, which are efficient upon excitation with x = 3650 A, were prepared by firing in HsS at 1200°C and quick cooling to room temperature. The emission spectrum depends on the In concentration and consists of three bands at 6200 A, 5350 A and 4700 A. X-ray analysis shows that the maximum amount of In incorporated is about 1O-2 gram-atoms In per mole ZnS. Refiring the phosphors at 600°C kills the luminescence at room temperature, an effect shown to be due to association. The trivalent In is compensated in the lattice bv Vi,‘. On the basis of the exoerimental evidence the 62OO,,A band is ascribed to In &,, the 5350 A b&d to the associate Ina~l&, and the 4700 A band to I’,,). * The low-temperature phosphorescence of the ZnS-In phosphors is strongly stimulated by irradiation into bandsat 1%~ and il.2~. 1. INTRODUCTION
IN THEIR work on the effect of trivalent substituents on the luminescence of Cu- and Agactivated ZnS KR&EE and DIKHOFF(~) found that several of the trivalent substituents introduced new emission bands. In the case of hexagonal ZnS-Ag,In they found a broad extra emission band with a peak at 6500 A. In hexagonal ZnS-Cu,In the peak of the extra emission was at 6800 A. From the fact that these bands, in contrast to the blue and green Ag- and Cu-emission, do not move on incorporation of Cd, KRUEGERand DIKHOFF concluded that the long-wave emissions could be ascribed to a characteristic transition in the Ins+ ion. KLASENS(~) found the emission peak of In in ZnS-Ag,In at 6100 A at -183°C and attributed it to the recombination of an electron in the In trap with a hole in the valence band. Because, as HOOGENSTRAATEN@) has shown, the distance between the trap level and the filled band does not change much on introduction of Cd, the noncharacteristic transition proposed by KLASENS also accounts for the fact that the In emission is little affected by Cd substitution. * For an explanation of the symbols used see Section 7. 69
FROELICH@), working with ZnS-Cu,In at relatively high In concentrations, found a weak longwave emission with a peak at about 6700 A at room temperature. The emission became brighter and shifted towards shorter wavelengths when the phosphor was cooled to - 190°C. Recently APPLE(~) investigated the system ZnS-x (Cu,Ag,Au), y (Ga,In) withy (added) > X. In hexagonal ZnS he found a Cu,In band peaking at 6400 A, an Ag,In band at 6140 A and an Au,In band at 7200 A. The characteristics of these emissions resembled those which appear in ZnS-Cu and ZnS-Ag without added coactivator and at relatively high Cu and Ag concentrations. In a later article APPLE and WILLIAMS@), assuming a strong association of activator and coactivator, ascribed the long-wave emissions found with In and Ga to a radiative transition from the ground state of the electron-trap to the ground state of the activator. In Section 10 the work of APPLE and of APPLE and WILLIAMS will be discussed in more detail and in relation to the results obtained in this work. The experiments described in this paper were carried out to get more information about the In emissions in ZnS by studying phosphors activated with In, without other added impurities. As will appear from the following sections ZnS phosphors
70
H. KOELMANS
activated with rather high concentrations of In may show a strong luminescence when cooled down quickly from the firing temperature to below 400°C. 2. EXPERIMENTAL PROCEDURE The samples were made from luminescent grade ZnS (Cl-free, precipitated from ZnS04). The concentration of In added ranged from 0.2 x 1O-2 to 2.0 x 10e2 gram-atoms In per mole ZnS. The In was added to the samples either by adding an In-solution for preparations with low In concentration or by adding the ternary compound
S- In
FIG. 1. Quartz
apparatus used in firing the phosphors.
ZnS-In
ZnInaS4 for preparations with high In concentration. It was checked that both ways of adding the In gave the same results. The powders were fired for 30 min at 1200°C in Oa-free HaS in quartz tubes placed in quartz vessels, as shown in Fig. 1. At 1200°C the InsSs is volatile as appeared from the deposition of InaSs on the cooler parts of the quartz vessel. In order to minimize the evaporation of InaSs the quartz tubes were closed with quartz caps during the firing (Fig. 1). It was established by chemical analysis that after firing, with about 10-a gram-atoms In per mole ZnS added, roughly 50 per cent of the added In remained in the samples. The In concentration found by chemical analysis gives, however, no clue about the amount of In really incorporated into the ZnS lattice. The amount of incorporated In was calculated from the change of the lattice constants of ZnS as is discussed in Section 3. After firing, the samples were quickly cooled to room temperature by taking the
quartz vessel from the furnace and dipping it into water. The powders showed a bright luminescence with 3650 A excitation in a colour ranging from bluish-green at low In concentrations to orangeyellow at high In concentrations. The light output of preparations with medium In concentrations was about 80 per cent of the output of commercial (Zn,Cd)S-Ag phosphors of comparable spectral distribution. The light output drops both at high and at low In concentrations. The light output at room temperature with 0*4,0*6, 1.2 and 4-O x 10-a In (added) is 40, 80, 100 and 45, respectively (output at 1.2x 10-s In put equal to 100). The phosphors showed only a weak emission on excitation with cathode rays. The rapid cooling of the powders after firing proved to be an essential part of the procedure. If the powders were allowed to cool slowly almost no luminescence was observed at room temperature. This phenomenon which strongly points to association effects is discussed in some detail in Section 8. Firing in sulphur vapour and firing in nitrogen gave substantially the same results. Apparently the phosphors are not very sensitive to changes in sulphur pressure. 3. THE AMOUNT OF In INCORPORATED INTO THE ZnS. High-angle X-ray powder diffraction patterns showed that the introduction of In produced a definite increase of the lattice spacings of ZnS. The results for different In concentrations (added before firing) are shown in Table 1. The c and a values of Table 1 were obtained by averaging values calculated from twelve different reflections. The concentrations of incorporated In as calculated from the increase in lattice spacings of the ZnS are shown in the last column of Table 1. In this calculation only the effect of substituting the bigger In atom for the smaller Zn atom was taken into account. The influence of the vacancies compensating the charge of the Ins+ was neglected. Comparison with solid solutions of Gas& in ZnS, which can be prepared at low temperature, with a wide range of Ga concentrations(‘) (so that the concentration of incorporated Ga is known from the preparation) shows that this approximation is not too bad. The Zn-S distance in hexagonal ZnS is 2.35 A. The In-S distance was taken to be 2.53 A, this being the In-S distance for In in the
ASSOCIATION
AND
DISSOCIATION
Table In cont. added, (gram-atom/mole)
1. Lattice
OF
constants of ZnS-In
c
4.0 x 10-s 1.2 x 10-s 0.4 x 10-s zero
6.2641 6.2624 6.2609 6.2606
f + + +
CENTRES
a.
0.0004 0.0004 0.0004 0.0004
3.8249 3.8238 3.8229 3.8223
+ + + +
0.0003 0*0003 0.0003 0.0003
IN
LUMINESCENT
ZnS-In
71
phosphors In cont. incorporated (gram-atom/mole)
0.8 x 1O-2 + 0.1 x 10-s 0.4 x 10-s * 0.1 x 10-s 0.1 x 10-s + 0.1 x 10-s zero
-
FIG. 2.
Excitation spectra of the ZnS-In different In concentrations.
tetrahedral arrangement in InaSa as taken from the work of HAHN and KLINGER(*). 4. EXCITA’I‘lON
AND
EMISSION
SPECTRA
The &S-In phosphors, especially those with high In concentrations, show a yellow body colour. The excitation spectra of the luminescence for different In concentrations are given in Fig. 2, the curves showing a maximum at about 4000 A
phosphors
with
for low and medium In concentrations, At high In concentration the region of strong excitation extends still further into the visible. The emission spectra for various In concentrations at room temperature and at - 190°C with h = 3650 A excitation are shown in Figs. 3(a) and 3(b). The emission bands are broad and of a composite nature. The maximum of the emission moves to shorter wavelengths with decreasing In
72
H.
KOELMANS
100
50
0 100
50
I lurn. A
O
I ! I 50
0
0 f-5
7.6 I
I
l-9
2-o
/
/
2-t /
2.2 I
?I00 680063aa 6aoll 5900 5sm
2.3
2.4 I
s-300
eV /
5000
I
4700
I
4400 a’
FIG. 3(a)
concentrations and with decreasing temperature. By analysing the emission curves it appeared possible to plot the over-all emission curves, both at room temperature and at -190°C for different In concentrations, as the sum of three normally
shaped sub-bands. As is usual in ZnS phosphors the emission (sub-) bands do not move significantly with temperature. The apparent shift of the overall emission curves with the In concentration and with temperature is entirely accounted for by
ASSOCIATION
AND
‘I.5
1.6
DISSOCIATION
I.7
I.8
,
I
I.9
2-O
I
/
OF CENTRES
2.1
2.2
2-3
24 /
7f55 6855 6555 6.205 5955 5650 5355
IN LUMINESCENT
2.5 j
5550
2.6
2.7 i
675u
2.8
2.9eV
/
Lm3
FIG, 3(b)
FIG. 3. Emission spectra with 3650 a excitation of the &S-In phosphors with different In concentration. (a) At room temperature. (b) At -19OY.L
a”
ZnS-In
73
H. KOELMANS
74
relative changes in the intensity of the sub-bands. The maxima of the sub-bands are at 6200 A (orange), 5350 A (green) and 4700 A (blue), respectively. The room-temperature emission at the highest In concentration is almost purely the orange band, whereas the - 190°C emission curve for the lowest In concentration essentially consists of the blue sub-band. Even though the proposed division into sub-bands accounts in a simple and consistent way for the observed emission spectra, such a division might seem somewhat arbitrary if we could not present arguments to support the analysis given. The low-temperature phosphorescence and the glow of the phoshors is of a different colour than the luminescence. This phenomenon is very striking in the sample with the lowest In concentration. The low-temperature luminescence is definitely blue whereas the lowtemperature phosphorescence and the glow are in the orange band. The low-temperature phosphorescence can be strongly stimulated with infrared (cf. Section 5); the emission is then again in the orange band. Furthermore the orange band at 6200 A closely corresponds to the orange emission in ZnS-Ag, In(s) whereas the same blue band at 4700 A is found in ZnS-Ga phosphors.@) A discussion of the emission bands and of the possible nature of the centres connected to them will be given in Section 9. 5. INFRA-RED
STIMULATION PHORESCENCE
OF
THE
PHOS-
The measurements were made with a Perkin Elmer 12C monochromator equipped with a quartz prism. The samples were cooled down to - 190°C and excited for 3 set with 3650 A radiation. After a waiting period of 70 set the samples were exposed to the infra-red. The infra-red stimulation spectrum of the lowtemperature phosphorescence (6200 A band) in a sample with Ini,e. = 0.8 x 10-s is given in Fig. 4. The stimulated light was detected with an R.C.A. type I.P. 21 photomultiplier tube. The curve of Fig. 4 represents the intensity of the emitted radiation for the same number of quanta at each wavelength and shows a weak band at about 0.7 eV followed by a strong band above 1.1 eV. Owing to measuring difficulties the maximum of this band could not be measured, but it probably extends
well into the visible. The spectral distribution of the emitted light was photographed on an Ilford panchromatic F.P. Special film. Excitation in the bands near 0.7 eV and above 1.1 eV only stimulated the orange band.
I 25wo
I,
I
I
~ 15mJ
~
10000
lOti
12000
, ~
a”
9030
FIG. 4. Infra-red excitation spectra of the low-temperature phosphorescence of ZnS-In phosphors.
6. TEMPERATURE EMISSION:
DEPENDENCE OF THE GLOW MEASUREMJZNTS
Figure 5 gives the temperature dependence of the total emission in a sample with Ini,c. = 0.4 x 10-s. It is seen that already at room temperature the emission is killed by about 40 per cent. Refiring the phosphor at 600°C (Fig. 5) in HsS causes the curve to shift towards lower temperatures, the room-temperature emission being almost completely quenched in this case (cf. Section 8). In the low-temperature emission of phosphors which were refired at 600°C the blue band was found to be absent. A discussion of this point will be given in Section 9. The results of the glow measurements (heating rate 2 degrees/set) are given in Fig. 6. The glow curve shows two peaks at -75°C and - 16O”C, respectively, the same peaks being found over the whole range of In concentration which
ASSOCIATION
AND
DISSOCIATION
OF
CENTRES
was investigated. At both peaks the glow is orange in all cases whereas the luminescence changes from bluish-green at low In concentrations to orangeyellow at high In concentrations.
IN
LUMINESCENT
ZnS-In
75
With this way of compensation In and S are incorporated in a 1 :l ratio and it is promoted by reducing conditions (low sulphur pressure). (b) [In&l = 2 [6&l With this way of compensation In and S are incorporated in a 2:3 ratio and it is promoted by neutral conditions. (4
II&J = PGkll With this way of compensation the In and S are incorporated in a 1:2 ratio and it is promoted by oxidizing conditions (high sulphur pressure).
In the equations above and in the following discussions the atomic notation of KRUGER and is used, viz ;
TI°Cj FIG. 5. Temperature dependence of the luminescence
v;,
=
of a ZnS-In phosphor fired at 1200°C and quickly cooled, and of the same phosphor refired at 600°C.
V’Zn = V7 Zn = n [V&-j =, I I
0
I
I
I
-150
*c
-100 4------
-50
FIG. 6. Glow peaks in ZnS-In 7. THE
MJXHANISM
I I
I
-200
phosphors.
OF INCORPORATXON OF In.
In view of the fact that the In atom is bigger than the Zn atom incorporation of In on the interstitial sites of the ZnS lattice is unlikely. So, assuming substitutional incorporation of In, according to K&GER and VINK@~), we have in principle the three ways of charge compensation mentioned below.
An In atom which has replaced a Zn atom; effective charge zero, denoted by 0. An In atom which has replaced a Zn atom; the In atom has lost one electron. Effective charge + 1, denoted by * . A Zn atom vacancy which has captured two electrons. This centre corresponds to a Zn ion vacancy in the ionic description. Effective charge - 2, denoted by ‘I. A Zn atom vacancy which has captured one electron. Effective charge - 1, denoted by ‘. A Zn atom vacancy. Effective charge zero. A free electron. Concentration of Zn atom vacancies etc.
Of the three ways of charge compensation mentioned above mechanism (a) can be ruled out because it would cause the phosphors to show n-conductivity, which is not the case. mechanism (c) cannot be operating as the main mechanism of charge compensation either, because then one would fmd free sulphur on dissolving the powders in acids, which is not the case. Even at the highest concentration of incorporated In, viz. about 10-z, no trace of sulphur is found on dissolving the samples in HCl. Thus, at least in the range of In concentrations investigated, only mechanism (b) remains. In the ionic picture this mechanism corresponds to the replacement of 3 Zns+ ions by
76
H.
KOELMANS
2 Ins+ ions, one Zn ion vacancy being formed. There are several striking examples of this kind of charge compensation in ZnS structures. The ternary compound ZnAlsS4, for instance, has, according to HAHN and FRANKLY),a ZnS (wurtzite) structure above 1000°C ; this means that one quarter of the cation-lattice sites are empty. An even more striking example of this kind is furnished by GasSa which, again according to HAHN and FRANKLY),may have a ZnS structure with one third of the cation sites unoccupied. It seems that in the ZnS-type of structure the close-packed arrangement of the anions is very stable, changes in the number and nature of the cations in the tetrahedral interstices being possible which do not affect the anion sublattice. 8. ASSOCIATION-DISSOCIATION EQUILIBRIA IN ZnS-In As was already stated in Section 2 the ZnS-In powders in order to show luminescence have to be cooled quickly from the firing temperature to room-temperature. If the phosphors prepared in this way are reheated at 500 to 600°C in HsS for as short a time as 5 min the room-temperature luminescence disappears almost completely. A refiring at temperatures below 500°C does not affect the luminescence. The killing of the fluorescence due to refiring at 600°C was found to be no function of the sulphur pressure used during refiring. When powders in which the room-temperature luminescence is destroyed by heating at 600°C are reheated at higher temperatures the room-temperature luminescence is gradually restored as the refiring temperature increases. The behaviour described here suggests a gradual association taking place when the samples are slowly cooled from 1200°C to 600°C. Below 500°C the motion of the centres through the lattice is frozen and for this reason quenched samples refired below 500°C do not change their behaviour. X-ray powder diffraction patterns of the powders refired at 600°C show that the lattice spacings, which have increased due to incorporation of In at 1200°C (cf. Section 3) practically drop* to those of pure ZnS in these samples. It thus appears that the high-temperature solution * The c and a values of phosphors refired at 600°C are 6.2610 A and 34229 A, respectively, as compared to 6.2606 A and 34223 A for the pure ZnS.
of Ins& in ZnS gets supersaturated on cooling, as a result of which a strong clustering of the dissolved In occurs. In the foregoing section it was argued that ;he compensation mechanism was [In&] = 2 [ Vz,]. It is plausible that the In&, and Vz, centres will attract each other. In the first place these centres have effective charges of differing sign. Further, because the In atoms are bigger than the Zn atoms which they replace, an trrangement in which the In& centres have a Vz, at a nearestneighbour site will be energetically more favourable than one in which In&, and Vz, are far apart (volume compensation). The latter effect is certainly one of the effects responsible for InsSs dissolving to a much smaller extent into ZnS than GasSa (cf. Section 3), the Ga atom having approximately the same size as the Zn atom. For the above reasons the onset of association may be represented by the following reactions: In’ + v” -+
InV’ with K1 =
[In’] [InVj
[v”] _
and InV’ + In’ --+
[In*] [InV’] InzVe with Kz = _____ [In2~“l
For convenience the symbols In&, Vi, etc are written In’,V’, etc. The reaction-constants Kl and Ks may be written in terms of association energies El and Ez. (cf. for instance LIDIARD(13)). By choosing reasonable values for Er and Ez the concentrations appearing in (a) and (b) may be calculated as a function of the amount of In incorporated and as a function of temperature. We will not give the calculation because in the forthcoming discussion we only need the general trends, which will be clear without a detailed calculation. These general trends are: (1) With increasing concentration of Inr,c. the concentrations of In’ and InV’ and especially the concentration of InsVe increase stronger than the concentration of V’. (2) On slowly cooling down the samples from high temperature the concentration of InsVa increases, whereas the concentrations of the other centres, and especially the concentration of V”, decrease.
ASSOCIATION
AND
DISSOCIATION
OF
In the next section we will try to correlate the centres In’, InV’, V” and InaVa with the three emission bands observed. 9. THE
NATURE
OF THE Z&-Ill
CENTRES
IN
The emission spectrum of &S-In phosphors was found to consist of bands at 6200 A, 5350 A and 4700 A (Section 4). Also the presence of the centres In’, InV’, V” and InaVe was made plausible. It is suggested that the following correlation exists between the emission bands and the centres : (a) The 6200 A band is due to recombination of electrons in the In traps and holes in the filled band, cf. KLASENS@). (b) The 5350 A band is due to the InV’ complexes. (c) The 4700 A band is due to the I”’ centres. (d) The InaP and bigger complexes act as killer centres. Below, arguments in favour of the assumed correlations will be given. The level scheme proposed for the ZnS-In phosphors is depicted in Fig. 7. On excitation, electrons are brought from Conductm .
bond .
. 53508
470061 1
2
3
4
7
h-6200X 5
"+yI,
-97 _
'"z""zn~-
8 ____ A
6
t --w
v
T
Valence band
FIG. 7. Level scheme of the ZnS-In phosphors. Levels occupied by an electron are indicated by a fully drawn line, empty levels by a broken line.
the V” and Inv centres to the conduction band, (processes 1 and 3 of Fig. 7). Some of these electrons drop back into the empty V” and InV centres (processes 2 and 4) thus giving rise to the blue and green emission bands, others fill the In traps (process 8). The orange band arises from the recombination of electrons from the In traps with holes in the valence band (process 5) originating from the empty V” and InV’ centres. It is likely that the orange band can also be excited
CENTRES
IN
LUMINESCENT
ZnS-In
77
directly (process 6). Without direct excitation the occurrence of a strong orange band (especially at high In concentration (cf. Fig. 3b) at -190°C would be hard to explain, because the escape of holes from the I”’ and InV’ centres is largely blocked at -190°C. The assumption that part of the light emitted in the orange band comes from direct excitation is further supported by the fact that excitation of the phosphor with 4358 A, instead of with 3650 A, was found to promote the orange band. The infra-red stimulation of phosphorescence and the glow are attributed to optical and thermal release of holes from the empty V” and InV’ centres (processes 9 and 10 in Fig. 7) towards the valence band followed by recombination with electrons trapped in the In traps, thus giving the orange band. The emptying of the hole traps is already operative at temperatures at which the release of electrons from the In traps (process 7) towards the conduction band is still small. So it is not surprising that in the glow experiments no indication of the well-known deep In trap is found. The position of the levels in Fig. 7 makes it logical to correlate the glow peak at -75°C to the InV centres and the glow peak at - 160°C to the V” centres. A further argument for this can be obtained from the consideration of the glow behaviour of comparable ZnS-Ga phosphors.@) These phosphors also show two glow peaks. One of them is also at - 160” (V”) the other at - 105°C (GaV’). The infra-red excitation peak of the phosphorescence > 1.1 eV is probably correlated to the Ink” centres and the band at about 0.7 eV to the V” centres. The assumption that InaYs and higher complexes act as killers, i.e. as centres which give way to a strong radiationless band-band recombination is corroborated by the fact that both a high In concentration and refiring at 600°C shifts the temperature dependence curve of the phosphors to lower temperature, a phenomenon typical for killer action. The influence of the killers will be the stronger the easier the excited centres release their holes. This means that the green emission and especially the blue emission will be quenched when the concentrations of InaVe and higher complexes rise. The orange emission, however, as far as it is directly excited by process 6 (Fig. 7), will be much less quenched because the electron trap due to In is much deeper than the hole traps of the empty
78
H. KOELMANS
InV’ and V” centres. The changes in relative importance of the emission bands with increasing In concentration as shown in Figs, 3(a) and 3(b) can now be explained qualitatively as follows:
likely. Firstly the blue band is quite strong; further the concentration of vz,, in ;he region of main compensation [I&J = 2 [V,J is proportional to the square root of the sulphur pressure. If the blue band were due tv V&, one might (a) The cont~bution of the blue band diminishes expect variations in intensity of the blue band on with increasing Ininca because the relative CORchanging the sulphur pressure, but, as was already cen~ation of V” rises less with increasing Intnc, stated in Section 2, the luminescence is not sensitive than that of the other centres (cf. Section 8), and to changes in sulphur pressure. further because it is more strongly quenched by the There are indications in the literature that the increasing concentration of InzV. blue bands abtained in ZnS with Cl, Al, Ga or by (b) The relative importance of the green band firing at high temperature may not all be at first increases with increasing .Inlnc. because the exactly the same plac&@ and may have a different concentration of InV’ rises, and then decreases shape.@” The maxima are found to vary between because it is more strongly affected by the in44OOk and 47OOk(15) Probably the self-activated creasing Inet’” concentrations than the orange blue emission bands are of a composite nature emission. as has also been suggested by other authors.Pl (cf The relative importance of the orange band The long-wave part (4700 A) of the self-activated increases with increasing Inme. because the orange emission could be due to V& and the short-wave emission is less quenched by the increasing conpart (4400 & to another centw, If this were true, centration of InsVs than the green and the blue this other centre would probably not be Vi,, emissions. because then the Fermi level would be between the The increased relative importance of the shorterlevels of V&., and V&,, i.e. close to the valence wavelength emission at low temperature (cf. band. The phosphors of this type should show Fig. 3b) is explained by a decreased migration of p-conductivity, which has never been observed. holes and consequently a decreased killer action In the level scheme of Fig. 7 the Jevel of the and a decreased transfer of the excitation energy effectively double-ahead centre, Vzn, was put to the centres with a longer-wavelength emission. at a smaller distance from the valence band than As has been stated in Section 6 the low-temperathe level af the XnZnV&, centre with its single ture emission spectra of phosphors which have negative charge. If the perturbation potentiak been refired at &@“C do not show the blue band. due to these centres were of a pure Coulomb This effect is very striking at low In concentrations. nature the relative positions of the levels should At a low temperature the luminescence in these be just the reverse. samples without the 600°C treatment is blue There are, however, more examples to show whereas the low-tem~rature luminescence of the that a Coulomb model does not hold. The atoms samples is yellow after refiring at 600°C. This Ni, Co and 0, for instance, which have the same phenomenon can be explained on the basis of the valence as the Zn or S atoms they replace, neveranalysis given. Due to refiring at 600°C the contheless give much deeper electron traps,@) than centration of Minsk” and higher complexes rises, Al vr Cl which have a valence different from the whereas especially the concentration of Yr is lattice constituents. strongly reduced. Both effects have a quenching In the present paper the long-wave emission at influence on the blue emission. 6200~%, found in ZnS-V&,In and ZnS-Ag,In, was connected to a radiative transition from the %@.coNcX.fULjPKJ REMARKS trap level to the valence band. APPLE and W~~L~A~S~%~ (cf. Section l), however, ascribed the fn Section 7 arguments were given fo; the cvmlong-wave bands they found with Ga and In to pensation mechanism [In&J = 2 [V&j in the radiative transitions from the ground state of the ZnS-In phvsphors discussed. It might be argued trap level to the ground state of the activator. in that the luminescence in, e.g. the 4700 If. band, the model of APPLE and WILLIAMS the charactercould nevertheless be due to a small concentration istics of the long-wave emissians found with a of another centre, e.g. [V&J. However, this is not
ASSOCIATION
AND
DISSOCIATION
OF
CENTRES
IN
Table 2. Long-wave en&ions in ZnS phosphors with In.
LUMINESCENT
ZnS-In
79
coactivated
Maximum of fluorescence Phosphor
ZnS-V”, ZnS-Ag, ZnS-Ag,
In In In
ZnS-Cu, In
Author
present work present work
at 20°C A
at -196°C A
6200 6200
6200 6200
6150 APPLEand WILLIAMS APPLEand WILLIAMS(~) 6350
given coactivator should depend on the chemical identity of the activator. In the present model this should not be the case. Table 2 gives a survey of our data about the long-wave emissions obtained with In and those of APPLE and WILLIAMS. It is seen from Table 2 that, in agreement with our model, the position of the long-wave emission does not depend on the nature of the activator for the combinations V”,In and Ag,In. The different location of,the peak with Cu,In, however, seems to favour the model of APPLE and WILLIAMS. There are, however, indications (also in APPLE’S own work@)) that at least part of the long-wave emission obtained with Cu,In is due to another centre, which could be identical with the red Cu-centre found by FROELICH(~) in non-coactivated ZnS-Cu phosphors. This might also explain the otherwise somewhat curious shift of the long-wave Cu,In emission to longer wavelengths (cf. Table 2), which occurs at low temperature. In our opinion the different position of the long-wave emission obtained with Cu,In cannot be used as a safe argument in favour of the model of APPLE and WILLIAMS. Summarizing the present situation one might say that the results with In seem to favour the idea of a long-wave emission characteristic for the coactivator, more work being necessary to elucidate the nature of the long-wave emissions obtained in ZnS activated with Cu alone and in ZnS activated with Cu combined with trivalent coactivators.
Acknowledgements-The
6150 6650
author
is indebted to Mr. G.
MEYER for measuring the infra-red-stimulated phosphorescence and to Dr. A. BRIL under whose supervision the excitation and emission spectra were measured. Thanks are due to Mr. LANG-IS for measuring the lattice constants of the &S-In phosphors. Thanks are also due to Dr. G. DIEMER,Mr. W. VANGOOL,Dr. W. HOOGENSTRAATEN and Dr. H. A. KLA~ENSfor their helpful criticism. REFERENCES 1. KRUGERF. A. and DIKHOFFJ., Physica s’Grav. 16,
297 (1950).
H. A., J. Electrochem. Sot. 100. 72 (1953). 2. KLASENS W., J. Electrochem.. SOC. 100, 3. HOOGENSTR~UTEN 356 (1953). 4. FROEL& H., J. Electrochem. Sot. 100. 496 (1953). Ele&ochem; So;. 5. APPLE E. F., Enlarged Abstr. 32, 45 (1958). 6. APPLE E.‘ F. and WILLIAMSF. E., J. Electrochem. Sot. 106,224(1959). 7. HAHNH. and co-workers, 2. anorg. Chem. 279, 241 (1955). 8. HAHNH. and KLINGERW., 2. anorg. Chem. 260, 97 (1949). H., unpublishedresults. 9. KOELMANS 10. KRUGER F. A. and VINK H. J., Solid State Phys. 3, 307 (1956). 11. HAHNH. and FRANKG., 2. anorg. Chem. 269, 227 (1952). 12. HAHNH. and FRANKG., 2. anorg. Chem. 278, 333 (1955). 13. LI&RD A. B., Phys. Rev. 94, 29 (1954). J. S. and WEIL D. J., Enlarged Abstr. Electro14. PRENER them. Sot. 32, 56 (1958). 1s. BUBER. H., J. Chem. Pkys. 20, 710 (1952); LEVEF+ ENZH. W., Luminescence of Solids, p. 188. John Wiley, New York (19SO).